Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!apple!motcsd!xdos!doug
From: doug@xdos.UUCP (Doug Merritt)
Newsgroups: comp.sys.amiga.tech
Subject: Re: setjmp/longjmp (Was: M2/C discussion)
Message-ID: <416@xdos.UUCP>
Date: 6 Jul 89 03:21:52 GMT
References: <4524@crash.cts.com> <18213@usc.edu> <8673@pyr.gatech.EDU> <413@xdos.UUCP> <26034@agate.BERKELEY.EDU>
Reply-To: doug@xdos.UUCP (Doug Merritt)
Organization: Hunter Systems, Mountain View CA (Silicon Valley)
Lines: 32

In article <26034@agate.BERKELEY.EDU> mwm@eris.berkeley.edu (Mike (I'll think of something yet) Meyer) writes:
> [ setjmp/longjmp ] are in the February 20th, 1987 draft.

Good.

>BTW, what makes you think the committee is only paying attention to
>general purpose CPUs? "noalias" was added at the behest of people doing
>implementations for vector processors.

I thought that was non-negotiable? :-)

Well, you see:
	1) If I hadn't qualified it, somebody would've said "There's this
	   hybrid APL/Lisp processor that can't longjmp because..."
	2) I imagine the committee cares the *most* about the most common
	   uses, so the point was that they wouldn't leave it out just
	   because some vector processor had trouble with it; it's too
	   valuable for the normal cases. Adding "noalias" (unesthetic
	   though it may be) is not quite the same as deleting a normal,
	   heavily used feature. (Ok, now raise the issue of constant
	   strings :-) But that's not *heavily* used.)
	Doug

>In any case, setjmp/longjmp have changed slightly. I vaguely recall
>hearing that it was at the behest of people doing C compilers for
>LISPMs.

Hmmm. Does this undermine me?
	Doug
-- 
Doug Merritt		{pyramid,apple}!xdos!doug
Member, Crusaders for a Better Tomorrow		Professional Wildeyed Visionary